American Journal of Epidemiology Vol. 154, No. 10 : 972-973
Copyright © 2001 by The Johns Hopkins University School of Hygiene and Public Health
BOOK REVIEWS |
Meta-Analysis, Decision Analysis, and Cost-Effectiveness Analysis: Methods for Quantitative Synthesis in Medicine, Second Edition
Division of Public Health Biology and Epidemiology School of Public Health University of California, Berkeley Berkeley, CA 94720
Only an epidemiologist living and working in a cave (without an Internet connection) could miss the massive explosion in both the number of meta-analyses being published (91 titles using the term "meta-analysis" reported in MEDLINE in 1990 and 395 in 2000) and textbooks devoted to the technique of meta-analysis (54 texts available for purchase at Amazon.com in July 2001). Dr. Diana Petitti's update of the earlier edition of this text is an able guidebook to this topic for both the novice and the intermediate meta-analyst. Along with the growth in meta-analysis has been a concurrent and similar increase in the use of the closely related tools of decision analysis and cost-effectiveness analysis, the two other methods treated in this book.
I have used the text as a primary source in several courses and seminars. Students at various levels of training regularly agreed that it was clearly written and provided numerous relevant examples of the necessary techniques. Often, methodological textbooks provide detailed formulae for the reader but leave it to the reader to actually "operationalize" the approach after deciphering the formulae. In contrast, this text provides in its numerous examples the actual step-by-step calculations that are needed. This allows the student reader to double check results that are "plugged into" a computer as well as to use the examples as a template for their own analyses. Although I think the text is most appropriate for students who have completed introductory (1-year) sequences in both epidemiology and biostatistics, I have no doubt that many of them could profitably use the book even prior to completion of these courses.
The book addresses the major topics and controversies in meta-analysis, decision analysis, and cost-effectiveness analysis. It opens with several examples of the applications as well as an overview of the techniques to motivate the remainder of the text. Early chapters provide excellent suggestions on areas common to all three techniques, including planning for these studies, the administrative approvals that might be required, and practical approaches to data collection. The section on information retrieval is particularly strong, with its quantitative demonstrations of the limitations of electronic search strategies and its examples of obvious publication biases related to nonpublication or inclusion of non-English articles and publication of only studies with positive significant findings. These sections have broad and obvious applicability to all research that uses previously published literature, not just to meta-, decision, and cost-effectiveness analyses.
The middle portion of the book sequentially reviews each of the three topics individually, including the necessary statistical methods. The clarity of this material (with its detailed examples of the calculations) is one of the book's great strengths and, in my experience, is not seen in many of the other textbooks on these topics. The chapter "Exploring Heterogeneity," in particular, helps to make clear for students what is often a very difficult and frustrating topic: how to evaluate and make sense of the data when the component studies of a meta-analysis have heterogenous results and why this heterogeneity may itself be a valuable finding.
There are, in my opinion, two minor limitations to the book. The author acknowledges in the preface that the text does not make much use of the Internet. There are, however, so many resources (good and bad) available on the Internet for students of these methods that a brief listing of the best web sites would be immediately useful to readers and be an appropriate complement to the textual material. There is also a wide variety of software packages available for implementing these techniques, and the reader would certainly appreciate even a brief comment from the author on her views of their relative strengths and weaknesses.
NOTES
ISBN 0–19–513364–1, Oxford University Press, New York, New York (Telephone: 212–726–6000, Fax: 212–726–6443), 2000, 306 pp., hardback $49.95
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||